Covariance calculation for batch processing terrain referenced navigation

Abstract

This paper proposes the method to calculate a covariance matrix of a batch processing terrain referenced navigation (TRN) for an integrated INS/TRN system using a Kalman filter. The batch processing TRN system cannot automatically provide a covariance matrix of the estimation error. That is why it is important to calculate a covariance matrix for batch processing TRN system. The newly adaptive algorithm based on recursive least square algorithm is proposed. The proposed adaptive algorithm can calculate the covariance of each measurement noise of the batch processing TRN. It also has stable property when the filter operates in a non-stationary environment. It is shown that the batch processing TRN with proposed algorithm has better performance than the TRN with conventional one in the computer simulations.